Will AI "replace" or "enhance" procurement?

The developments over the last year in Generative AI are difficult to ignore. However, with any new technology, questions inevitably arise about its impact on, and sometimes even the elimination of, roles in our field.

In this article, I will explore a few ideas from my time at Waymo (Google's self driving-car program) to help answer this question.

When people first see a LightSource demo, we're often met with a flurry of different reactions:

Wow this is incredible sci-fi tech!

Is this going to replace my job?

How can I learn to take-advantage better of AI tools, when the field is so new to me and changing quickly?

These are all natural concerns for anyone to have. However, to spoil the ending: I strongly believe that, in the long term, new technology will serve to enhance the capabilities of procurement professionals, not replace them.

During my tenure at Waymo , we were developing an alternative to Uber with vehicles capable of self-driving. Concurrently, there was a semi-truck program. People often asked me "what else could be automated?"

I once met an entrepreneur with an idea for an Autonomous Cargo Ship. The concept involved equipping a container freight vessel with a range of sensors and computers, enabling it to cross the Pacific and dock at ports without a single human onboard.

Is this a good idea? No, it's a terrible one.

Often procurement is managing hundreds-of-millions or even billions of spend. Replacing those jobs (instead of enhancing procurement with better tools) would be foolish.

Consider a cargo ship loaded with freight, carrying approximately $150M worth of goods. The average ship is crewed by 20-30 people, with smaller vessels staffed by as few as 8. The cost of labor is a very small price to pay relative to the risk of failure.

Let's do the math:

20x people x 15-day trip x $120k per year salary =

~$100k of labor per voyage.

This calculation shows that labor costs represent only 0.07% of the value of the cargo on the ship alone.

Therefore, rather than replacing the humans on the ship, a better solution is to empower them with improved sensors, data, and control. This enhances their ability to perform more effectively, given the significant value at stake.

The same principle applies to procurement. Replacing jobs with automation, instead of providing procurement professionals with better tools, would be counterproductive.

The potential impact, or financial loss, from not enhancing procurement practices far outweighs the cost of personnel and the necessary technology. In this context, CFOs should be the most significant advocates for funding procurement innovation.

However, the equation reverses when considering small-dollar 'tail spend.' In these situations, where managing the spend is too time-consuming, automation becomes a more sensible option. This is akin to ride-sharing services like Uber and Lyft, where the 1:1 ratio of driver to passenger makes the ROI of autonomy high.

What are your thoughts? I'm eager to hear in the comments below — are there areas where you think AI will indeed replace procurement jobs?

Special thanks to Sergio Sernani, FMVA for our inspiring conversation earlier that helped brainstorm and form these thoughts!

Spencer Penn is the CEO and co-founder of LightSource , a leading procurement technology company based in San Francisco. Spencer formerly helped lead the Model 3 program at Tesla and is regularly featured in major conferences and press outlets on topics ranging from Electrification, to Autonomy, AI, and Supply Chain.